Basic Econometrics Final Exam (Trimester 1/2003)

BEcon Program, Faculty of Economics, Chulalongkorn University

Instructions:
a) Textbooks, lecture notes and calculators are allowed.
b) Each must work alone. Cheating will not be tolerated.
c) There are four(4) tests. Attempt all the tests. Each carries equal weights.
d) Use only the provided test-books.
e) All the hypothesis testing will use 0.05 as the level of significance.
TEST#1 (20 points)
An opinion survey of general households about government-run insurance program has been
conducted to evaluate the proposed “Eau-arthorn” insurance scheme. The following model has
been used:
Prob(BUYi=1) = (Zi) ---------------(1)
Zi = 1 + 2 FSIZi + 3 INCi/FSIZi
where BUYi = 1 if household i decided to buy
= 0 ,otherwise
(.) = cumulative distribution function (CDF) of Standard Normal
FSIZi = size of household i (persons)
INCi = income of household i (thousand baht per year)
EViews printout for the model estimation using Maximum Likelihood method is shown below
============================================================
Dependent Variable: BUY
Method: ML - Binary Probit
Date: 09/20/03 Time: 22:54
Sample: 1 100
Included observations: 100
Convergence achieved after 5 iterations
Covariance matrix computed using second derivatives
============================================================
Variable Coefficient Std. Error z-Statistic Prob.
============================================================
C -2.725997 1.149618 -2.371220 0.0177
FSIZ 0.152895 0.146313 1.044980 0.2960
INC/FSIZ 0.042802 0.012158 3.520532 0.0004
============================================================
Mean dependent var 0.630000 S.D. dependent var 0.485237
S.E. of regression 0.385506 Akaike info criteri 0.908186
Sum squared resid 14.41564 Schwarz criterion 0.986341
Log likelihood -42.40928 Hannan-Quinn criter 0.939816
Restr. log likelihood -65.89557 Avg. log likelihood -0.424093
LR statistic (2 df) 46.97258 McFadden R-squared 0.356417
Probability(LR stat) 6.31E-11
============================================================
Obs with Dep=0 37 Total obs 100
Obs with Dep=1 63

September 23, 2003-Assoc. Prof. Pongsa Pornchaiwiseskul Page

1/6
Basic Econometrics Final Exam (Trimester 1/2003)
BEcon Program, Faculty of Economics, Chulalongkorn University
============================================================
Coefficient Covariance Matrix
================================================
C FSIZ INC/FSIZ
================================================
C 1.321621 -0.156520 -0.012859
FSIZ -0.156520 0.021408 0.001326
INC/FSIZ -0.012859 0.001326 0.000148

Based on the above printout, answer the following questions:

1.1) Explain why or why not the given printout is appropriate for the estimation of model
(1).
1.2) if 100 households with four (4) persons and annual income of 100 thousand baht are to
be asked for their opinion, how many households will you expect to buy the insurance?
1.3) If a household in question 1.1 is to have one more family member next year, given the
same income level, test the hypothesis that the probability of the household’s buying
insurance will be unchanged. Explain in details.
Hint: H0: 42 + (100/4)3 = 52 + (100/5)3
TEST#2 (20 points)
Foreign direct investment and stock market return have been postulated to have lagged own
and cross effects. Their rate of change can be expressed in VAR(1) as follows:
FDIt = 1 FDIt-1 + 2 RMt-1 + u1t ------------------(2.1)
RMt = 1 FDIt-1+ 2 RMt-1 + u2t ------------------(2.2)
where FDIt = rate of change of foreign direct investment in period t
RMt = rate of change of stock market return in period t
u1t,u2t = error terms or shocks for FDI and RM in period t,
respectively
Given the following printouts,

September 23, 2003-Assoc. Prof. Pongsa Pornchaiwiseskul Page

2/6
Basic Econometrics Final Exam (Trimester 1/2003)
BEcon Program, Faculty of Economics, Chulalongkorn University

September 23, 2003-Assoc. Prof. Pongsa Pornchaiwiseskul Page

3/6
Basic Econometrics Final Exam (Trimester 1/2003)
BEcon Program, Faculty of Economics, Chulalongkorn University

answer the following questions.

2.1) Do estimate equations (2.1) and (2.2). That is, estimate 1,2,1,2 var(u1t) and var(u2t)
2.2) Assume simultaneous shocks, determine the weight of shock series (u1t and u1,t+1) in
total variation of FDIt+1 and RMt+1

September 23, 2003-Assoc. Prof. Pongsa Pornchaiwiseskul Page

4/6
Basic Econometrics Final Exam (Trimester 1/2003)
BEcon Program, Faculty of Economics, Chulalongkorn University
Hint: Sources of total variation of FDIt+1 and RMt+1 are (u1t ,u2t) and (u1,t+1, u2,t+1)
2.3) Check the validity of your estimates in 2.1). Is there any long-term relationship between
FDIt and RMt? Explain.
2.4) Given that FDIt and RMt have a long-term relationship (regardless of your answer in
question 2.3), what will be your answer to question 2.2?
TEST#3 (20 points)
Return of Food sector during period t has been postulated to follow equation (3.1) below:
RFD t – RGOV t = 1 +2(RSET t – RGOV t) + u t-1 + t ----------------(3.1)
where RFDt = return of food sector in period t
RGOVt = interest rate of government bond in period t
RSETt = average return of stock market in period t
ut = error term in period t
ut = u t-1 + t
t = white noise in period t with V(t)=2 for all t
Estimation of Equation (3.1) based on 80 periods of time series data is shown below:

September 23, 2003-Assoc. Prof. Pongsa Pornchaiwiseskul Page

5/6
Basic Econometrics Final Exam (Trimester 1/2003)
BEcon Program, Faculty of Economics, Chulalongkorn University

Based on the above EViews printout, answer the following questions:

3.1) Given RFD80 =0.02, RGOV80 =0.01 and RSET80 =0.15
and RSET81 =0.18 and RGOV81=0.011, predict the return of food sector in period t=81.
3.2) Provide the prediction interval for the answer in question 3.1
3.3) Test whether the error term u is stationary. Explain in details.
TEST#4 (20 points)
Back to TEST#3, if the following model is used instead of model (3.1)
RFDt – RGOVt = 1 +2 (RSETt – RGOVt) + u t ----------------(4)
2
Var(ut) = 0 + 1Var(ut-1)+ 2(ut-1)
where ut’s are assumed to be stationary and independent error terms, explain how you will
answer questions 3.1and 3.2 in this case.

End of Exam.
Good luck.

September 23, 2003-Assoc. Prof. Pongsa Pornchaiwiseskul Page

6/6